The long-term objective of the present proposal is to uncover novel targets for treating malignant glioblastoma. This proposal explores a novel interaction between the PI3-K signaling cascade involving PTEN, and a group of pro-metastatic proteins, Ezrin-Radixin-Moesin (ERM), in regulating the motile behavior of glioblastoma. Glioblastoma multiforme is the most malignant form of human glioma characterized by extensive local invasion. Identification of molecular targets to treat this malignancy will have an impact on the clinical outcomes. Several oncogenes have been shown to be overexpressed in glioblastoma. These include Epidermal Growth Factor Receptor (EGFR), Platelet-derived Growth Factor (PDGF), and the receptors for PDGF. In late stage glioblastoma, the most frequent genetic alteration is the loss of PTEN tumor suppressor. The molecular mechanisms responsible for the highly invasive nature of glioblastoma have not been fully elucidated. The major finding is that PTEN mediates the disassembly of motile pseudopodia in glioblastoma. This is correlated with a drastic redistribution of a membrane-cytoskeleton linker protein, moesin, and a concomitant reduction in cell motility. The regions within PTEN responsible for this activity are the phosphatase domain and the last four amino acid PSD95/Dlg/ZO-1 (PDZ)-binding domain. Interestingly, PDZ-binding domain is specifically targeted for mutation inactivation in glioblastoma. This application will develop a novel brain tumor model, Pten?PDZ, which harbors a deletion in the PTEN PDZ-binding domain to recapitulate the invasive phase of glioblastoma development. In Specific Aim 1, primary astrocytes will be prepared from both wild-type and Pten?PDZ-/- mice. The signaling cascades linking PTEN to Akt and ERM will be delineated. The activation states and subcellular distribution of ERM, PTEN, and Akt in these astrocytes will also be compared. By using ectopic expression and siRNA, the role of Akt in regulating ERM proteins will be determined. In Specific Aim 2, the migratory properties of astrocytes prepared from Pten?PDZ+/+ and Pten?PDZ-/- mice will be compared. The role of Akt and ERM proteins in cell migration will be examined by silencing RNA strategy. In Specific Aim 3, Pten?PDZ-/- mice will be crossed to a RasB8 glioma model. Gliomas generated in PasB8: Pten?PDZ+/+ and RasB8: Pten?PDZ-/- mice will be analyzed for their histological grades, invasive features and relative tumor latency. These studies will lay the foundation for the development of molecular-based therapeutics for brain tumors. PUBLIC HEALTH RELEVANCE: Human glioblastoma multiforme is a highly malignant tumor and the clinical prognosis is generally poor. This proposal explores the use of a novel preclinical mouse model to recapitulate the invasive phase of this malignant disease. The long-term goal is to use this mouse model as a drug discovery tool to identify compounds that can block the spread of human glioblastoma.